There is one simple target for Gordon Brown and David Cameron (and, in theory, Nick Clegg) on 6 May. It is to secure at least 326 seats in the next House of Commons. With 650 MPs to be elected, that is the number needed to secure an overall majority of two.
This target has increased a little since the last election. After a review of constituency boundaries, four new seats are being created in England, so the target is up by two seats on the 324 required in 2005. The review, which has taken place everywhere apart from Scotland (the number of seats there was cut from 72 to 59 at the last election), makes assessing the parties' prospects in each seat a little harder. Unless a seat has been unaltered by the review, we have to rely on estimates of what would have happened in each new constituency in 2005 if it had been in place then.
Those estimates, constructed by Colin Rallings and Michael Thrasher of Plymouth University, have been put together carefully. They use local election results to gauge the impact of the changes of geography in each seat. But there is inevitably room for disagreement. And where voters find themselves voting in a constituency that is very different politically from what they are used to, the new tactical situation might induce them to vote differently.
The new boundaries are intended to ensure that the constituency map catches up with population changes across Britain. In general, our cities and the more urban areas are either losing population or growing more slowly than suburban and rural areas. That means that the populations, and electorates, in Conservative seats are growing more quickly than in Labour ones. Correcting the resulting imbalance in electorate sizes, the new boundaries make it somewhat tougher for Labour to win this time. According to the Plymouth University estimates, if the new boundaries had been in place in 2005 Labour would have had an overall majority of 48, rather than the 66-seat majority it actually secured. The Liberal Democrats' overall tally of 62 seats would have been unchanged.
But if the boundary review makes life a bit tougher for Labour, it is still likely to be easier for Labour than the Conservatives to reach the magic total of 326 seats. This is because the Tories find it more difficult to turn votes into seats. One reason is that the boundary review is already seriously out of date. In England it was undertaken on the basis of population data that is now at least 10 years old. And the review has also done nothing about the deliberate over-representation of Wales and the (now) unintended over-representation of Scotland. Overall, on 6 May around 4,000 more people will be eligible to vote in the average Tory seat than in the average Labour one. Two further factors add to the imbalance. First, fewer people typically turn out to vote in safe Labour seats than in safe Tory ones. Second, the Conservative vote is inefficiently distributed: as things stand, in a close election the Conservatives will win too few marginal seats by narrowly beating Labour. The resulting "bias" against the Conservatives is illustrated in the table, which shows what would happen for various national divisions of the vote on the assumption that each division occurred as a result of the Labour vote going down and the Tory vote going up by exactly the same proportion in each constituency. It assumes too that the Liberal Democrats, along with all the other smaller parties, do as well as they did in 2005. On these assumptions, Gordon Brown will be able to reach the target of 326 simply by winning the same share of the vote as the Conservatives. In contrast, the Conservatives will have to be nearly 11 percentage points ahead to secure an overall majority. They will have to be more than five points ahead just to win more seats than Labour.
What these figures also show is that there is a wide range of results that could produce a hung parliament. That range will narrow somewhat if the Liberal Democrat vote falls, but not by a great deal. For example, if the Liberal Democrats fall back five points to 18 per cent, the range still extends from the Conservatives being ahead by three-quarters of a point (which would still give Labour a majority) to their being ahead by just under nine percentage points.
This wide range helps explain the considerable speculation about the possibility of a hung parliament. In many constituencies the Liberal Democrats have developed strong roots locally and are therefore much more effective than before in turning votes into seats, a development that inevitably makes hung parliaments more likely. But inevitably, the party (if any) that hits the target of 326 seats will be determined above all by what happens in the crucial marginals. If the Conservatives can secure a bigger swing to themselves in these seats, they should be able to reduce some of the bias against them, and thereby increase their chances of hitting the target. The Tories need to add 116 seats to the notional tally of 210 seats which the Plymouth University team estimates they would have won in 2005. We can define these 116 seats as the ones the Tories are projected to capture to reach 326 if the assumptions underpinning the calculations in our table are fulfilled. Fifteen of them are seats the Tories would hope to capture from the Liberal Democrats or the SNP as a swing to the Tories from Labour would enable the party to leapfrog locally into first place.
However, the vast majority – 101 of the 116 seats the Tories need to gain – are currently held by Labour. They are concentrated in three crucial "hotspots", where the election will effectively be won or lost. These hotspots are the North-west and Yorkshire; the Midlands; and London and the South-east. Each contains roughly a quarter of the key Labour-Tory battlegrounds. Scotland and Wales contain few such hotspots.
The typical marginal seat is a place where a largely working-class, urban Britain meets a more middle-class, rural Britain. Many such seats are part of the commuter towns and suburbs that encircle many of our major cities. In the North they include Bury North, Bolton West and Batley & Spen. In the Midlands, Dudley North, Redditch and Warwick are typical. In and around the capital they include constituencies such as suburban Croydon Central and Harrow East together with commuting towns such Crawley and Dartford.
The three battlegrounds are connected by the M1/M6 corridor. It is a motorway route with which our politicians are going to become very familiar in the coming weeks.
John Curtice is Professor of Politics at Strathclyde University